This subdivision of the Mediterranean into separate basins, divided from each other by shoals or submarine ridges, by islands and promontories, sufficiently explains the contrasts between the phenomena of the open ocean and those observed here. In the Mediterranean, it is well known, the tides are almost everywhere irregular and uncertain. To the east of the Narrows of Gibraltar, in the sea extending between Andalusia and Morocco, the tides are hardly felt at all, and {26} they are, moreover, interfered with to such an extent by currents that it is exceedingly difficult to determine their amplitude, or the establishment of the various ports. Nevertheless the rise and fall of the tidal wave are sufficiently marked to have attracted the attention of Greek and Italian navigators. On the coasts of Catalonia, France, Liguria, Naples, Asia Minor, Syria, and Egypt the oscillation is hardly perceptible, but on those of Eastern Sicily and of the Adriatic the tide sometimes rises three feet, and, if accompanied by storms, may even attain a height of ten feet in certain localities. The Straits of Messina and of Euripo (Eubœa) have their regular tides, and in the Gulf of Gabes the waters rise and fall with the same regularity as in the open ocean. In the Black Sea, however, no tidal movements whatever have been discovered hitherto. It is nevertheless probable that more careful observations will lead to the discovery of a feeble tide, for it is believed that this phenomenon exists even on Lake Michigan, which has only one-fifth the area of the Black Sea. [Illustration: Fig. 6.—THE STRAIT OF GIBRALTAR. According to Robiquet, Randegger, and others. Scale 1 : 750,000. Erhard.] The Mediterranean differs not only from the open ocean with respect to the feebleness and irregularity of its tides, but it is likewise without a great stream-current keeping in constant circulation the whole body of its waters. The currents which have been observed in various divisions of the Mediterranean can be ascribed only to local causes. An Italian geographer of the last century, Montanari, has {27} advanced an hypothesis of a great circuit current which entered the Mediterranean through the Strait of Gibraltar, and, after having washed the shores of Africa as far as Egypt, returned to the west along those of Asia and Europe; but careful observers have vainly endeavoured to discover its existence. They have met only with local currents, produced by an indraught of the waters of the Atlantic, by winds, by the floods of rivers, or by an excess of evaporation. One of these currents sets along the coasts of Morocco and Algeria from west to east; another flows along the Italian coast of the Adriatic from north to south; and a third from the mouth of the Rhone in the direction of Cette and Port Vendres. In fact, the configuration of the sea-bottom, and particularly the shoal between Sicily and Tunis, precludes the existence of any but surface currents in the Mediterranean.